Week 7: Agenda power: who decides what the public is asked about

In plain words

Week 7 of the My Vote Forward seminar studies agenda power: the power to decide what gets asked, discussed, and decided at all. Students read Bachrach and Baratz’s 1962 “Two Faces of Power,” McCombs and Shaw’s 1972 study of how news coverage shapes which issues voters think are important, and recent research on social media feeds and local newspaper closures.

Any list of priorities has to start with a list. Whoever writes the list has already shaped the answers, because people cannot rank a subject that is not there. News coverage and social media also shape which subjects feel urgent. This week asks how that power works and how it can be made visible.

Defined terms

• Agenda setting: the process by which some issues come to be seen as important and others do not. See agenda-setting.
• Salience: how prominent or important an issue is to a person or the public at a given time. See salience.
• Nondecision: in Bachrach and Baratz’s account, the suppression or exclusion of an issue so it never reaches a decision.
• Framing: the way an issue is described, which shapes which considerations people bring to it.
• Agenda capture: control of what a body considers by a narrow group. See the site’s chapter on agenda capture.

Learning objectives

By the end of week 7, a participant should be able to:

1. Explain Bachrach and Baratz’s “second face” of power and why it challenged pluralist community power studies.
2. State McCombs and Shaw’s agenda-setting hypothesis and what kind of evidence supports it.
3. Connect Zaller’s account of elite framing (week 6) to agenda setting.
4. Summarize what the 2023 Facebook and Instagram field experiments found, and did not find, about feeds and polarization.
5. Describe the Darr, Hitt, and Dunaway finding on newspaper closures and its limits.
6. Propose procedures that would make a priorities channel’s own agenda choices transparent and contestable.

Core readings

1. Peter Bachrach and Morton S. Baratz, “Two Faces of Power,” American Political Science Review 56(4), 1962, pp. 947 to 952.⁠¹
2. Maxwell McCombs and Donald L. Shaw, “The Agenda-Setting Function of Mass Media,” Public Opinion Quarterly 36(2), 1972, beginning p. 176.⁠²
3. John R. Zaller, The Nature and Origins of Mass Opinion (Cambridge University Press, 1992), chapter on elite discourse (review from week 6).⁠³
4. Andrew M. Guess and coauthors, “How Do Social Media Feed Algorithms Affect Attitudes and Behavior in an Election Campaign?” Science 381(6656), 2023, pp. 398 to 404.⁠⁴
5. Andrew M. Guess and coauthors, “Reshares on Social Media Amplify Political News but Do Not Detectably Affect Beliefs or Opinions,” Science 381(6656), 2023, pp. 404 to 408.⁠⁵
6. Joshua Darr, Matthew P. Hitt, and Johanna Dunaway, “Newspaper Closures Polarize Voting Behavior,” Journal of Communication 68(6), 2018, pp. 1007 to 1028.⁠⁶

Lecture notes

The question before the question

Agenda power is the subject of week 7, and it sits upstream of everything the course has studied so far. Week 5 showed that a group ranking depends on the counting method. Week 6 showed that answers depend on question wording and order. Week 7 adds a more basic dependence: results depend on which questions are asked at all. A priority list with ten subjects cannot reveal that the public cares most about an eleventh. The site’s objection page who chooses the ten? poses this challenge to My Vote Forward directly. This lecture gives it its theoretical and empirical background.

Bachrach and Baratz: the second face of power

Peter Bachrach and Morton Baratz wrote into a dispute about community power in the early 1960s. Their abstract describes the dispute: sociologically oriented researchers “consistently found that power is highly centralized,” while political scientists “just as regularly concluded that in ‘their’ communities power is widely diffused,” the latter styling themselves “pluralist” and their counterparts “elitist.” Bachrach and Baratz argue that “the pluralists themselves have not grasped the whole truth of the matter”; while their criticisms of the elitists are sound, both groups “utilize an approach and assumptions which predetermine their conclusions.” Their central thesis is “that there are two faces of power, neither of which the sociologists see and only one of which the political scientists see”.⁠¹

The face the pluralists saw was power in observable decisions: who wins when an issue is contested. The second face, as Bachrach and Baratz develop it in the article, is the power to keep issues from being contested at all, by shaping the values, procedures, and rules that determine what reaches decision. Studying only decisions misses the issues that never became decisions. Students should read the article for its full argument; the summary here is the seminar’s.

The relevance to between-election channels is immediate. A consultation that asks residents to choose among three options for a new road has already excluded the option of no road. A participatory budget that covers only capital projects has excluded operating budgets. A priority list that omits a subject has excluded it. None of these choices is visible in the results, which is exactly Bachrach and Baratz’s point.

McCombs and Shaw: the media set the agenda

Maxwell McCombs and Donald Shaw gave agenda setting its name and its first systematic test. Their abstract states the claim: “In choosing and displaying news, editors, newsroom staff, and broadcasters play an important part in shaping political reality. Readers learn not only about a given issue, but also how much importance to attach to that issue from the amount of information in a news story and its position.” In reflecting what candidates say, “the mass media may well determine the important issues, that is, the media may set the ‘agenda’ of the campaign”.⁠²

The agenda-setting hypothesis concerns salience, not persuasion. The media may not tell people what to think, but they influence what people think about. That distinction matters for priorities. A priority ranking measures salience, which issues people consider most important. If salience is shaped by coverage, a priority ranking partly measures what the media covered recently. That is not a reason to dismiss priority data. It is a reason to interpret it alongside a record of media attention, and to ask whether a ranking reflects durable concerns or the news cycle.

Zaller: elites frame the terms

Week 6 introduced Zaller’s premise that people answer with the considerations most salient to them. His book also emphasizes “the role of political elites in establishing the terms of political discourse in the mass media and the powerful effect of this framing of issues on the dynamics of mass opinion on any given issue over time”.⁠³ Taken with McCombs and Shaw, this produces a two-stage picture: elites and media influence which issues are salient, and then influence which considerations people bring to those issues. A priorities channel operates inside this picture. It cannot stand outside it.

New media: what the field experiments found

The public debate about social media often assumes that platform algorithms drive polarization and set the political agenda. Two 2023 field experiments in Science tested parts of this assumption during the 2020 United States election, with consenting users of Facebook and Instagram.

In the first, Andrew Guess and coauthors assigned users to reverse-chronological feeds instead of the default algorithms. Moving users out of algorithmic feeds “substantially decreased the time they spent on the platforms and their activity.” It also changed what they saw: “the amount of political and untrustworthy content they saw increased on both platforms,” uncivil content decreased on Facebook, and content from moderate friends and ideologically mixed sources increased on Facebook. Yet “despite these substantial changes in users’ on-platform experience, the chronological feed did not significantly alter levels of issue polarization, affective polarization, political knowledge, or other key attitudes during the 3-month study period”.⁠⁴

In the second, the researchers removed reshared content from users’ feeds for three months. This “substantially decreases the amount of political news, including content from untrustworthy sources,” decreased clicks and reactions, and “reduces partisan news clicks.” It also “produces clear decreases in news knowledge within the sample.” But “contrary to expectations, the treatment does not significantly affect political polarization or any measure of individual-level political attitudes”.⁠⁵

These results deserve careful reading from every direction. They do not show that social media has no effect on politics: the studies ran for three months, during one election, with consenting users, changing one feature at a time. They do show that two widely proposed changes substantially altered exposure without detectably changing attitudes over that period. People who argue that platforms drive polarization and people who argue they do not should both engage with these findings rather than citing only the parts that fit. The site’s chapters on misinformation and polarization discuss the broader evidence.

Local news: the disappearing agenda

The decline of local newspapers is a different kind of agenda change: not a new force shaping attention, but the loss of an old one. Joshua Darr, Matthew Hitt, and Johanna Dunaway argue that “as local newspapers close, Americans rely more heavily on available national news or partisan heuristics to make political decisions.” Using genetic matching “to compare counties that are statistically similar but for the loss of a local newspaper,” they “identify a small but significant causal decrease in split-ticket voting in presidential and senatorial elections”: “in areas where a newspaper closed, split-ticket voting decreased by 1.9%”.⁠⁶

For this course, the finding suggests that local subjects may lose salience when local coverage disappears, and national frames may fill the gap. A priorities channel for a city or county would then measure priorities shaped partly by national media, which may not match local responsibilities. The site’s chapter on local news collapse develops the point, and the subject pages’ “who decides” chapters, for example who decides on housing, map which level of government holds which powers.

Making agenda power visible

No priorities channel can avoid agenda power. Someone must choose the subjects, their labels, and their order. The question is whether those choices are visible and contestable. Several procedures follow from the readings, and the seminar presents them as design proposals, not established practice:

• Publish the process for choosing subjects, including who chose them and what was considered and rejected.
• Offer an open “other” option and report what people write in, so excluded subjects can surface.
• Randomize the order of subjects for each participant.
• Report media attention to each subject during the collection period alongside the ranking.
• Revisit the list on a published schedule, with a documented way to propose changes.

These steps do not remove agenda power. They move it into the open, which is the most that Bachrach and Baratz’s analysis allows.

A worked example: the missing subject

Hypothetical example: a county publishes a priority survey with eight subjects: roads, parks, policing, schools, libraries, water, transit, and economic development. Housing is left out because the county has limited authority over it. The survey finds roads ranked first. A resident group argues that housing would have ranked first had it been listed, and that its absence shaped the results. The county replies that listing housing would have misled residents about what the county can do.

Both arguments have force, and neither can be tested from the published results alone. This is the structure Bachrach and Baratz describe: the decision that mattered most was made before anyone ranked anything. An open “other” option would have produced some evidence, at the cost of inviting organized write-in campaigns. A note explaining the exclusion and pointing to who decides on housing would have made the agenda choice contestable without pretending it was neutral. The example also shows why the subject list’s authority boundaries matter: a priority list that mixes subjects a body controls with subjects it does not can mislead participants about what their ranking could change.

Discussion questions

1. Think of a public decision in your community. What options were never on the table? Who decided that, and how?
2. McCombs and Shaw distinguish salience from persuasion. Is a priority ranking a measure of public concern, or a measure of recent media attention? How could you tell?
3. The 2023 feed experiments changed exposure without detectably changing attitudes in three months. What would you need to see to conclude that platforms do or do not shape political attitudes over longer periods?
4. If local news decline shifts attention to national issues, should a local priorities channel correct for that? Would correcting for it be a form of agenda power?
5. Who should choose the subjects on a public priority list: officials, experts, random citizens, or participants themselves? What does each choice risk?
6. Is an open “other” option enough to counter agenda power, or would it simply favor the organized?

Assignment

Agenda audit (1,500 words). Choose a real public consultation, survey, or participatory process with a published list of options or questions, such as a city budget survey or a participatory budgeting ballot from one of the site’s cases. Identify who set the agenda and how, what options or subjects were excluded, and whether any mechanism existed to add them. Apply Bachrach and Baratz to identify any nondecisions, and McCombs and Shaw to consider how media attention may have shaped the results. Propose two specific changes that would make the agenda more transparent, and state the strongest objection to each.

Evidence

Bachrach and Baratz, McCombs and Shaw, both Guess and coauthors studies, and Darr, Hitt, and Dunaway are quoted from their published abstracts. Zaller is quoted from the book’s published description. The description of the “second face” beyond the abstract is the seminar’s summary and is labeled as such. The procedures under “Making agenda power visible” are design proposals, not evidence of practice. All sources were opened on the review date.

Competing interpretations

Pluralists argued that power is best studied in observable decisions, and that claims about nondecisions are hard to test; Bachrach and Baratz replied that studying only decisions predetermines a pluralist conclusion. On media, some researchers emphasize strong agenda-setting effects, while others stress that audiences select and resist coverage. On social media, one view holds that algorithms and reshares drive polarization and misinformation; the 2023 field experiments found substantial changes in exposure without detectable changes in attitudes over three months, which supporters of the first view argue is too short and narrow a test. On local news, the Darr, Hitt, and Dunaway effect is causal by their design and small in size; readers differ on its importance.

Application to this framework

My Vote Forward’s priorities proposal asks people to rank a fixed list of subjects. This week shows that the list is itself an exercise of agenda power and that rankings will partly reflect media attention. The framework should be judged on whether it discloses who chooses the subjects, how they can be changed, and how results are interpreted against the news environment, as discussed in who chooses the ten? and agenda capture.

Unresolved questions

• What procedure for choosing a priority list would be accepted as fair across the political spectrum?
• Can priority rankings be adjusted for media attention without introducing a new, less visible agenda choice?
• Do longer-term exposure changes on platforms affect attitudes in ways three-month experiments cannot detect?